Editorials
developing countries it is important to assess in which
patients the benefit from being given magnesium is
sufficient to justify this risk. Treatment is certainly justi-
fied in women with eclampsia, in whom evidence from
meta-analysis indicates that magnesium reduces
mortality. A quarter of the women in the Magpie study
had severe pre-eclampsia—very high blood pressure
( > 170 mm Hg systolic or 110 mm Hg diastolic) with
very high proteinuria, or lower blood pressure (150
mm Hg systolic or 100 mm Hg diastolic) with two or
more signs of imminent eclampsia such as hyper-
reflexia, frontal headache, blurred vision, or epigastric
tenderness). In this group it was necessary to treat 63
women to prevent one seizure. In women who did not
have such severe pre-eclampsia 109 patients had to be
treated to prevent a seizure. Even the women without
severe pre-eclampsia were probably quite ill in this
study, as almost 75% of them were given antihyperten-
sive treatment. Thus, the Magpie study indicates a very
favourable ratio of benefit to risk for magnesium, given
according to the protocol, in women with severe
pre-eclampsia or requiring antihypertensive treatment.
The safety of magnesium in this study was
facilitated by limiting the loading dose of magnesium
to 4 g and restricting intravenous administration to
1 g/hour, whereas the intramuscular dose was at 10 g,
followed by 5 g every 4 hours. For the loading and
intravenous doses this is considerable lower than has
been recommended by some, and the safety of higher
doses is not assured by this study.4 In addition, some
instruction was undoubtedly provided to the partici-
pants in the trial. None the less, as carried out in this
protocol with simple clinical assessment and without
determining magnesium concentration, treatment with
magnesium was safe.
Despite the evidence, this effective treatment has
not been used widely. We have few examples in obstet-
ric practice of treatments that have been tested in ran-
domised controlled trials to show efficacy and even
fewer that address treatment in the field. Why has this
Average length of stay, delayed discharge, and
hospital congestion
A combination of medical and managerial skills is needed to solve the problem
T
he NHS is under sustained pressure to cope
with rising numbers of hospital admissions.
Public concern over waiting on trolleys and
delays in access to care has never been greater. The
NHS has responded by using its most expensive
resource—inpatient beds—more efficiently. Over the
past 20 years the average length of stay for each admis-
sion has fallen year on year from 11.7 days in 1980 to
6.8 in 1999-2000. Factors have included increased use
of day surgery and the recognition that earlier
discharge in many conditions was not dangerous and
may often be better for the patient.
After nearly 20 years of consistent reductions, the
average length of stay has unexpectedly risen from
610
treatment not become part of the armamentarium of
providers of obstetric care throughout the world? The
answer is complex, but at least part of the explanation
is that this inexpensive generic treatment has no indus-
trial advocate to facilitate licensing, production, and
distribution. Another factor is the reluctance of care
providers and administrators to change healthcare
practice. On behalf of the World Health Organization,
Fédération Internationale de Gynécologie et
d’Obstétrique, and the International Society for the
Study of Hypertension in Pregnancy we advocate the
use of magnesium sulphate in the treatment and
prevention of eclampsia. We urge nations in which
eclampsia has a major impact on maternal mortality to
institute policies to ensure that this inexpensive and life
saving treatment is made available and that care
providers are trained to use it safely.
James M Roberts past president
International Society for the Study of Hypertension in Pregnancy,
Magee-Womens Research Institute, 204 Craft Avenue, Pittsburgh,
PA 15213, USA (rsijmr@mail.magee.edu)
Jose Villar coordinator
Maternal Health Research Department of Reproductive Health and
Research, World Health Organization, Geneva 27, Switzerland
(villarj@who.int)
Sabaratnam Arulkumaran treasurer
Fédération Internationale de Gynécologie et d'Obstétrique,
Department of Obstetrics & Gynaecology, St George’s Medical
School, London SW17 0RE (sarulkum@sghms.ac.uk)
Competing interests: None declared.
1 Eclampsia Trial Collaborative Group. Which anticonvulsant for women
with eclampsia? Evidence from the collaborative eclampsia trial. Lancet
1995;345:1455-63.
2 Duley L, Henderson-Smart D. Magnesium sulphate versus diazepam for
eclampsia. Cochrane Database Syst Rev 2000;(2):CD000127.
3 The Magpie Trial Collaboration Group. Do women with pre-eclampsia,
and their babies, benefit from magnesium sulphate? The Magpie Trial: a
randomised placebo-controlled trial. Lancet 2002;359:1877-90.
4 Sibai BM, Lipshitz J, Anderson GD, Dilts PV Jr. Reassessment of
intravenous MgSO4 therapy in preeclampsia-eclampsia. Obstet Gynecol
1981;57:199-202.
6.8 days in 1999-2000 to 6.95 days in 2000-1. The rise
shown in the national hospital episode statistics for
England may seem small in absolute terms, but a 2.5%
rise has huge potential costs at all levels of the service.
A recent workshop attended by professionals and
department of health officers examined the figures and
noted that the rise was apparent for both elective
admissions (1.0%) and non elective admissions (2.9%),
was present in all major adult specialties, and was
present in all regions of the country. The changes were
too consistent to be dismissed as chance. The largest
change (11.5%, in mental health) may be explained by
recent changes in service configuration between
community and hospital, but the 6.6% increase for BMJ 2002;325:610–1
BMJ VOLUME 325 21 SEPTEMBER 2002 bmj.com

general medicine (including geriatrics) has no obvious
explanation. Furthermore, analysis by age decade
shows that the rising average length of stay was present
in the over 85s for three years, in the over 75s for two
years, and in the over 65s this year. The challenge is to
find an explanation.
Delayed discharge has long been a concern.1 2
There is no agreed definition, so reliable data are
scarce, but studies show that delayed placement into
institutional care after completion of the assessment
and rehabilitative process is an important factor.3 4 So
too are the reductions in numbers of beds in nursing
homes, problems in funding from social service budg-
ets, and waits for assessments from therapists or social
services, for community services, or for equipment to
be ordered, delivered, and installed. Hospitals with
greater than 85% occupancy have no flexibility to cope
with this.5
But there are new factors too. Hospitals are complex
systems of health and social care6—action in one area
usually affects others and not always predictably. Waits
on trolleys in the emergency room become inevitable if
patients cannot be discharged. If medical teams are con-
centrated at the front end of the admission, then those
same staff cannot also perform the essential tasks of pre-
paring patients for discharge, often on inappropriate
wards. Reduced junior doctor hours and introduction of
shift working has resulted in less time on the wards to
see patients and their relatives, leading to inefficient
communication and poor operation of discharge proce-
dures. Integrated working of multidisciplinary teams will
have to rely less on junior doctors and may require the
development of physician assistants or other staff
working in discharge coordinator roles.7 8
Patients become acutely ill every day, and there
would seem to be opportunities to convert five day
radiological and laboratory departments to seven day
services. However, a national shortage of radiogra-
phers and legal constraints on who can use radiation
means that implementation of such plans is difficult.
For consultants (who have an average working
week of more than 60 hours9) to fill the gaps created by
reduced junior doctor hours cannot be done without
sacrificing other duties such as outpatient clinics, and
other targets such as the two week wait for patients with
cancer. Many of these pressures will get worse as the
working time directive bites harder in 2004.
There is no single solution. Intermediate care
remains an ill defined entity and some schemes have
simply reallocated beds currently used for rehabilita-
tion.10 Intermediate care could have an important role
but requires new resources, strong leadership, and
dedicated medical time to ensure that it is care of
equivalent quality in an alternative environment, not
simply cheaper care.
The perverse financial incentives for local social
services to ensure timely discharge must be tackled,
and perhaps the United Kingdom should consider
aligning the financial incentive with the needs of
patient care, as in Scandinavia.11 12 The patient’s
charter raised expectations of patients and their
relatives. Even when they are fit for discharge, patients
can elect to remain in hospital until a final placement is
found. Does society intend this or should there be legal
mechanisms to insist on interim placements?
BMJ VOLUME 325 21 SEPTEMBER 2002 bmj.com
Editorials
The national service framework for older people is
intended to improve not only the quality of patients’
experience but also the efficiency of services; it could
help by providing leadership and the right incentives.
Most of the answers to these very complex problems lie
in combined medical and managerial approaches at
local level that result in realistic and achievable recom-
mendations, within clear national planning. The
challenge is to ensure the framework and resources are
in place to allow such partnerships to function.
David Black consultant geriatrician
Queen Mary's Hospital, Sidcup, Kent DA14 6LT
Mike Pearson director, clinical effectiveness and evaluation
unit
Royal College of Physicians, London NW1 4LE
Competing interests: None declared.
1 Rubin SG, Davies GH. Bed blocking by elderly patients in
general-hospital wards. Age Ageing 1975;4:142-7.
2 Epstein J, Kaplan G, Lavi B, Noy S, Ben Shahar I, Shahaf P, et al. A
description of inappropriate hospitals stays in selective in-patient
services: a study of cases receiving social work services. Soc Work Health
Care 2001;32:43-65.
3 Koffman J, Fulop NJ, Pashley D, Coleman K. No way out: the delayed dis-
charge of elderly mentally ill acute and assessment patients in North and
South Thames Regions. Age Ageing 1996;25:268-72.
4 British Geriatrics Society. Three national surveys of delayed discharge.
London: BGS, 1996-8. (BGS newsletters, Sep 1996, Mar 1997, May 1998.)
5 Bagust A, Place M, Posnett JW. Dynamics of bed use in accommodating
emergency admissions: stochastic simulation model. BMJ
1999;319:155-8.
6 Black DA, Bowman C, Severs M. A systems approach to elderly care. Br J
Healthcare Manage 2000;6:49-52.
7 Bridges J, Meyer J, Davidson D, Harris J, Glynn M. Hospital discharge.
Smooth passage. Health Service J 1999;109:24-5.
8 Royal College of Physicians. Skill mix and the hospital doctor. New roles for
the healthcare workforce. London: RCP, 2001.
9 Royal College of Physicians. Workforce survey 2001. London: Royal
College of Physicians, 2001.
10 Implementation of the National Service Framework and Intermediate Care seen
from a Geriatricians’ and Older People’s Perspectives. London: Age Concern
and the British Geriatrics Society. 2002.
11 Health Services Journal. Councils could be forced to fund their
bed-blockers. Health Services Journal 2002;28th February: 4.
12 Styrborn K. Thorslund N. Bed blockers: delayed discharge of hospital
patients in a nationwide perspective in Sweden. Health Policy
1993;26:155-70.
Correction
Growth hormone in growth hormone deficiency
In this editorial by Paul Saenger (13 July, pp 58-9), we inad-
vertently mixed up some references and failed to print the
author’s competing interests. In the reference list, references
7, 8, 9, and 10 should be numbered 8, 9, 10, and 11 respec-
tively; reference 11 should not be there at all; and the new
reference 7 should read:
Schoenau E et al on behalf of the Germany Lilly Growth
Response Study Group. A new and accurate prediction
model for growth response to growth hormone treatment
in children with GH deficiency. Eur J Endocrinol 2001;
144:13-20.
The sentence “The mean duration of treatment was 6.2
years—the duration of treatment was thus twice as long as
the French study and the dose of treatment was also twice as
much, that is, 0.3 mg/kg/week (0.9 IU/kg/week compared
to 0.14 mg/kg/week),” towards the end of the penultimate
paragraph, should be supported by reference 8. Other
references numbers in the text remain the same.
Professor Saenger has advised the growth hormone
industry on scientific issues and has received honorariums.
611